RESEARCH Open Access

Implementation of next generationsequencing into pediatric hematology-oncology practice: moving beyondactionable alterationsJennifer A. Oberg1† , Julia L. Glade Bender1,6†, Maria Luisa Sulis1,6†, Danielle Pendrick1, Anthony N. Sireci2,Susan J. Hsiao2, Andrew T. Turk2, Filemon S. Dela Cruz1,6,7, Hanina Hibshoosh2,6, Helen Remotti2, Rebecca J. Zylber1,Jiuhong Pang2, Daniel Diolaiti1,7, Carrie Koval3, Stuart J. Andrews2, James H. Garvin1,6, Darrell J. Yamashiro1,2,6,Wendy K. Chung1,4,6, Stephen G. Emerson4,5,6, Peter L. Nagy2,8, Mahesh M. Mansukhani2,6 and Andrew L. Kung1,6,7*

Abstract

Background: Molecular characterization has the potential to advance the management of pediatric cancer andhigh-risk hematologic disease. The clinical integration of genome sequencing into standard clinical practice has beenlimited and the potential utility of genome sequencing to identify clinically impactful information beyond targetablealterations has been underestimated.

Methods: The Precision in Pediatric Sequencing (PIPseq) Program at Columbia University Medical Center institutedprospective clinical next generation sequencing (NGS) for pediatric cancer and hematologic disorders at risk for treatmentfailure. We performed cancer whole exome sequencing (WES) of patient-matched tumor-normal samples and RNAsequencing (RNA-seq) of tumor to identify sequence variants, fusion transcripts, relative gene expression, and copynumber variation (CNV). A directed cancer gene panel assay was used when sample adequacy was a concern.Constitutional WES of patients and parents was performed when a constitutionally encoded disease was suspected.Results were initially reviewed by a molecular pathologist and subsequently by a multi-disciplinary molecular tumor board.Clinical reports were issued to the ordering physician and posted to the patient’s electronic medical record.

Results: NGS was performed on tumor and/or normal tissue from 101 high-risk pediatric patients. Potentiallyactionable alterations were identified in 38% of patients, of which only 16% subsequently received matchedtherapy. In an additional 38% of patients, the genomic data provided clinically relevant information ofdiagnostic, prognostic, or pharmacogenomic significance. RNA-seq was clinically impactful in 37/65 patients(57%) providing diagnostic and/or prognostic information for 17 patients (26%) and identified therapeutictargets in 15 patients (23%). Known or likely pathogenic germline alterations were discovered in 18/90patients (20%) with 14% having germline alternations in cancer predisposition genes. American College ofMedical Genetics (ACMG) secondary findings were identified in six patients.(Continued on next page)

* Correspondence: kunga@mskcc.org†Equal contributors1Department of Pediatrics, Columbia University Medical Center, New York, NY10032, USA6Herbert Irving Comprehensive Cancer Center, Columbia University MedicalCenter, New York, NY 10032, USAFull list of author information is available at the end of the article

© The Author(s). 2016 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Oberg et al. Genome Medicine (2016) 8:133 DOI 10.1186/s13073-016-0389-6

(Continued from previous page)

Conclusions: Our results demonstrate the feasibility of incorporating clinical NGS into pediatric hematology-oncologypractice. Beyond the identification of actionable alterations, the ability to avoid ineffective/inappropriate therapies,make a definitive diagnosis, and identify pharmacogenomic modifiers is clinically impactful. Taking a more inclusiveview of potential clinical utility, 66% of cases tested through our program had clinically impactful findings and samplesinterrogated with both WES and RNA-seq resulted in data that impacted clinical decisions in 75% of cases.

Keywords: Whole exome sequencing, RNA sequencing, Precision medicine, Pediatric oncology

BackgroundThe outcomes for children with cancer have steadily im-proved to the present time when more than 80% of allpediatric oncology patients are cured [1]. Nonetheless,cancer remains the leading cause of disease-related death inchildren. Moreover, this success has come at a price; two-thirds of all survivors have some long-term sequelae attribut-able to their treatment [2]. Together, the requirement tofurther improve existing outcomes and to decrease toxicityunderscores the need for the current national initiative inprecision medicine to include pediatric oncology patients.Many of the advances in pediatric oncology have resulted

from the implementation of risk-stratified treatment strat-egies that incorporate histological, anatomical, and molecu-lar prognostic and predictive determinants into the choiceof therapies for individual patients [3]. Changes in ploidy,chromosomal segmental changes, and specific gene alter-ations are routinely utilized to guide treatment intensity inpediatric oncology [4]. Therefore, the tenants of precisionmedicine are intrinsic to the practice of pediatric oncology.Recent advances in massively parallel sequencing allow

for more comprehensive approaches to determine theabnormalities contributing to tumorigenesis. Initialimplementation of next generation sequencing (NGS)technologies focused on the identification of actionablealterations, with estimates in the range of 5% to nearly100% depending on disease histology and evolving defini-tions of “actionable” [5–15]. The utility of these technolo-gies, however, extends well beyond the identification ofactionable alterations and determining the value of thesetechnologies should be more inclusive and consider thebroad clinical impact of testing.In 2014, we implemented a clinical NGS platform for

pediatric oncology patients. The Precision in PediatricSequencing (PIPseq) Program utilizes NGS of tumorand germline in a CLIA-certified (Clinical LaboratoryImprovement Amendments of 1988) environment andincludes interrogation of both DNA and RNA. We con-ducted a retrospective review of our first 101 consecu-tively sequenced patients utilizing the PIPseq pipelineand report here our experience with integrating clinicalNGS into pediatric hematology-oncology practice anddescribe the broad clinical utility of genomically informedcancer medicine.

MethodsThe PIPseq pipelineTo achieve more comprehensive genome-level analysisin our pediatric oncology patients, we utilized threeCLIA-certified, CAP (College of American Pathologists),and New York State Department of Health approvedassays. When possible, we utilized a cancer whole exomesequencing test (cWES) comprising WES of tumor andnormal tissue (buccal swab or peripheral blood) andRNA sequencing (RNA-seq) of tumor tissue. This assaywas optimized for fresh or frozen specimens. Whensample adequacy was a concern, we also utilized a di-rected cancer gene panel assay which sequenced 467cancer-associated genes and was optimized for use withformalin fixed paraffin embedded (FFPE) material(Columbia Comprehensive Cancer Panel, CCCP). If aconstitutionally encoded disease was suspected (e.g.familial hemophagocytic lymphohistiocytosis), we per-formed constitutional WES from the patient and bothparents (trio) when available.Tissue for sequencing was obtained either from ar-

chived blocks (FFPE) or frozen tissue blocks from theDepartment of Pathology. DNA and RNA extraction andsequencing were performed in a CLIA-certified labora-tory. Variant calls were independently made on tumorand germline material and somatic variants determinedby subtraction. Copy number variation (CNV) was de-termined from the WES data, fusion transcripts wereidentified from the RNA-seq data, and relative gene ex-pression was determined by comparison to a model builtfrom 124 transcriptomes. A mix of tissues was used togenerate the model including normal white blood cells,lung, liver, brain, glioma, and cell lines.After initial review by a molecular pathologist, all re-

sults were reviewed in a multi-disciplinary moleculartumor board. Participants included representation bymolecular pathology, pediatric oncology, cytogenetics,medical genetics, and cancer biology. For each patient, areport was issued containing variant calls, CNV, fusions,and overexpressed genes. Variants were assigned a tierbased on disease-association and separately a tier basedon level of evidence for clinical actionability (describedbelow). Reports were delivered to ordering oncologistsand posted to the electronic medical record (EMR) in

Oberg et al. Genome Medicine (2016) 8:133 Page 2 of 19

accordance with patient opt-in/opt-out preferences se-lected at the time of informed consent.

Patients and informed consent for clinical sequencingBetween January 2014 and April 2016, NGS was per-formed on tumor and/or normal tissue from 101 high-riskpatients by the Laboratory of Personalized GenomicMedicine at Columbia University Medical Center(CUMC). This represented approximately 32% of the totalpatients in our clinical practice. High-risk patients weredefined as those having a prognosis of <50% overallsurvival at 5 years, outlier clinical phenotype, rare cancerwithout standard of care therapy, suspected cancerpredisposition, or relapsed disease. A request for constitu-tional WES, cWES and RNA-seq, or targeted cancer paneltesting was made at the discretion of the referring oncolo-gist in consultation with the PIPseq team [16].Participants signed consent for WES or cWES either

as part of an Institutional Review Board (IRB)-approvedprotocol (IRB nos. AAAB7109, AAAJ5811) or theysigned the clinical consent (http://pathology.columbia.edu/diagnostic/PGM/oncologytests.html). Written con-sent for clinical WES and cWES testing was obtainedafter the risks and benefits had been explained to thepatient and/or caregiver, which include the potentialdisclosure of medically actionable secondary findings,defined as germline disease-causing mutations unrelatedto the condition for which sequencing was being per-formed. Patients could opt in or opt out of the following:learning secondary findings and/or having these resultsappear in the EMR; having their samples and/or datastored for future research, both with or without identi-fiers; and future contact. Results not reported includedcarrier status, variants of uncertain significance (VOUS)in secondary findings except as related to cancer, andmutations related to adult-onset conditions for whichthe genetic link is either unclear or for which no knownintervention is of proven benefit (e.g. Alzheimer’s dis-ease). IRB approval was obtained for this retrospectiveanalysis of de-identified patient and clinical genomicsdata (IRB nos. AAAP1200 and AAAQ8170).

Clinical sequencingTesting required at least 200 ng of DNA for WES, at least50 ng of DNA for targeted DNA sequencing, and at least3000 ng of RNA for transcriptome analysis (Additional file1: DNA and RNA extraction). The entire assay was aCLIA-certified assay. The laboratory-developed test usedgeneral purpose reagents and the Agilent WES ver.5 +UTR baits. Specifically, WES was performed using theAgilent SureSelectXT All Exon V5 +UTRs capture kitfor library generation and sequenced on the HiSeq2500using paired-end 125 cycle × 2 sequencing (twotumors, two normal and two transcriptomes, pooled

together and run in two lanes). Targeted DNA sequen-cing was performed on a 5.59 Mb Custom AgilentSureSelectXT library, targeting 467 genes, and sequencedon a HiSeq2500 using paired-end 125 cycle × 2 sequencing(seven samples per lane). RNA was sequenced using theTruSeq Stranded Total RNA LT Sample Prep Kit with 125cycles × 2 paired-end sequencing on the HiSeq2500.

Sequencing analysisDNA sequencing reads were de-multiplexed and con-verted to fastq files using CASAVA from Illumina.Mapping and variant calling of tumor and normal sam-ples was performed using NextGene (v.2.3.4; Softge-netics, State College, PA, USA), which uses a modifiedBurrows-Wheeler transform (BWT) alignment method.Sequences were mapped to GRCh37 (“hg19”), retainingreads with a median quality score of 20 or higher, withno more than three ambiguous bases, a minimum num-ber of 25 called bases per read, and trimming readswhen three consecutive reads fell below a quality scoreof 16. Alignment and variant calling was performedusing paired-end reads with a minimum of 10 reads, atleast three variant reads, and a minimum variant allelicfraction of 10% for tumor and 5% for normal was re-quired to call a variant. The variant calling module wasset to “detect large indels.” The variant calling algo-rithm showed a 99.6% agreement with single nucleotidepolymorphisms on an oligonucleotide microarray andover 96% sensitivity in inter-laboratory comparison anda 96% detection rate for heterozygous variants in a 40/60% mixture of samples. For small indels, the lab de-tected 93% of all variants detected by another labora-tory in inter-laboratory comparison, with the greatestdisagreement in insertions greater than 10 bp.Variants were subject to filtering. In normal DNA,

variants were passed through a “reference range filter” ofcancer predisposition genes, genes relevant to pharma-cogenomics, and variants relevant to patient care; a“reportable range filter” which includes COSMIC(cosmic70 provided by Annovar) variants in the patient’smutation report file and variants in genes recommendedby the American College of Medical Genetics (ACMG)for the reporting of secondary findings [17]; as well as afrequency filter, which includes variants whose minorallele frequency in the 1000 Genomes (phase 1, version 3,release date 23 November 2010) is less than 1%. Somaticmutations in the tumor were identified by subtracting allvariants called in normal tissue (output at minor allelicfraction of ≥5%) from variants called in the tumor (outputat a minor allelic fraction of ≥10%). The approach maxi-mized the number of variants output to minimize thelikelihood of filtering out actionable mutations prior tomolecular tumor board discussion (Additional file 1:Supplementary Methods; Somatic Variant Calling Strategy).

Oberg et al. Genome Medicine (2016) 8:133 Page 3 of 19

Variants in the tumor were further characterized ashomozygous, compound heterozygous, somatic, and“disruptive” (loss of function, namely, nonsense, frame-shift, or splice site). Spreadsheets with the various cat-egories were presented to molecular pathologists for re-view. Quality statistics for WES and cWES are presentedin Additional file 2: Table S1. Targeted DNA sequencingwas performed to an average 500X depth and analyzedas above. All DNA sequencing results were manuallyreviewed by molecular pathologists to prioritize variantsfor presentation at the multi-disciplinary tumor boardand subsequent reporting of consensus variants. Formutation statistics, the list of “tumor-specific” variantsobtained by comparison of vcfs was filtered for variantswith at least 30X coverage in tumor and either a “qualityscore” ≥20 or a variant allelic fraction ≥25% in the tumor.

Copy number variationCNV was identified using EXCAVATOR (v.2.2; https://sourceforge.net/projects/excavatortool) software [18].For samples with greater than 95% of targeted nucleo-tides present at least 10X in the reference normal andat least 90% covered 30X in the corresponding tumorsample, EXCAVATOR was run with parameters chosenfor moderate sensitivity (assuming a tumor percentageof 0.8) and cutoff for loss set to a log2 ratio of –0.2. Inaddition, all high quality heterozygous variants with vari-ant allelic fractions (VAFs) in the range of 45–55% and90–100% in the normal sample were output. The allelicratio at these genomic coordinates in the tumor was alsooutput for viewing on the integrated genomic viewer toallow identification of copy number neutral loss ofheterozygosity (LOH) and to support the CNVs identifiedby EXCAVATOR. The laboratory detected all chromo-some arm changes seen on karyotyping, losses of 26 Mband greater seen on array CGH, and reproducibly identifiedall CNVs that involved at least ten exons at 40% tumorfraction (Additional file 1: Supplementary Methods).

Transcriptome analysisFor transcriptome analysis, fastq files from CASAVAwere filtered for ribosomal RNA (rRNA) using Sort-MeRNA (v.1.7; http://bioinfo.lifl.fr/RNA/sortmerna/)and trimmed to remove poor-quality tails using Trim-Galore (v.0.2.7; http://www.bioinformatics.babraham.ac.uk/projects/trim_galore/) with settings to exclude readsof quality score <20 and read length <20. Remainingreads were mapped to GRCh37 (hg19) using TuxedoSuite [19, 20] consisting of TopHat2 (v.2.0.8), BOW-TIE2 (v.2.1.0), and CUFFLINKS (v.2.1.1). Non-uniquelymapped reads were excluded before estimation offragments per kilobase per million reads (FPKMs) byCUFFLINKS. Mutation calling was performed usingNextGene software. At least 50 million uniquely

mapped reads were required with less than 5% DNAcontamination. In addition, unmapped reads were ana-lyzed using “FusionMap” (v.01/01/2015) to generate alist of fusions for review by molecular pathologists [21].To identify alterations in gene expression, the medianFPKMs of 8000 housekeeping genes were used as refer-ence [22] and the relative expression of each gene wascompared to 124 normal transcriptomes from varioustissues (13 blood, 20 liver, 24 kidney, 17 lung, and 50brain) (Additional file 1: Supplementary Methods).

Data interpretation and reportingInterpretation of clinical WES, RNA-seq, and CNV wasconducted via a molecular tumor board with multi-disciplinary representation from pediatric oncology, path-ology, cancer biology, molecular and clinical genetics, andbioinformatics. Following the tumor board, approximately60 days after testing request, a tiered report was generatedfor clinical samples by pathology, sent to the referringphysician and posted to the EMR according to the opt-in/opt-out selections of the patient consent. Only variantswith good normal coverage (generally at least 30X) weredetected on multiple independent fragments and were notexcluded as likely benign were reported. For clinical test-ing, the report included variants that were justified by theliterature as driver mutations (e.g. well characterized hotspot mutations); unambiguous loss of function mutationsin tumor suppressor genes (i.e. nonsense or frame-shiftmutations that resulted in loss of functional domains); mu-tations with published laboratory data documenting gainor loss of function in oncogenes and tumor suppressorgenes, respectively; and previously reported fusions or fu-sions that were expected to have the same effect as previ-ously reported fusions involving one of the partner genes.Certain exceptions for clinical testing were made. For ex-ample, if a variant was likely a strong driver (e.g. a known ac-tivating mutation of an oncogene) but had low coverage inthe normal or appeared low-quality on review, the molecularpathologist still considered it but required independentconfirmation by an orthogonal method prior to reporting.The final clinical cWES report included: known

tumor type-specific actionable somatic mutations (Tier1); somatic mutations in targetable pathways, actionablesomatic mutations in other tumor types, somatic muta-tions in well-established cancer genes (Tier 2); othersomatic mutations in cancer genes (Tier 3); and som-atic VOUS (Tier 4). Reporting of germline findingsincluded: known pathogenic secondary ACMG variants[17]; secondary non-ACMG variants and selectedVOUS in known cancer genes with commentary; andknown variants that influence pharmacogenomics.Reports further included translocations, significantlyoverexpressed genes, and segmental CNV. A samplecWES report is presented in Additional file 3. Accession

Oberg et al. Genome Medicine (2016) 8:133 Page 4 of 19

number for all genes and fusions referenced in the paperare reported in Additional file 2: Table S2. Datasets areavailable through the cBioPortal for Cancer Genomics(http://cbioportal.org) [23, 24].Clinical utility, defined as the ability of a molecular

test result to provide information related to the care ofthe patient and his/her family members to diagnose,monitor, prognosticate or predict disease progression,and inform treatment [25], was used to evaluate thepotential impact of findings from clinical sequencing.“Clinical impact” and “clinically impactful” are broadterms used throughout this paper to refer to any mo-lecular test result that, when integrated with a patient’shistory, symptoms, and other clinical findings, informedthe medical team’s assessment or management of thepatient. These clinically meaningful results were subca-tegorized into the following five categories to evaluatethe potential clinical utility of tumor and germline alter-ations: (1) diagnostic; (2) prognostic; (3) identification ofa therapeutic target; (4) other clinically impactful infor-mation, including pharmacogenomics or findings thatresulted in a significant refinement of a therapeutic plan(e.g. choice of donor or withdrawal of recommendationfor bone marrow transplantation); and (5) recommenda-tions for health maintenance interventions or geneticcounseling for the patient and other at-risk familymembers. Genetic alterations were considered targetableif: (1) an FDA approved drug or experimental drug wasavailable that inhibited the target directly or inhibitedits downstream signaling pathway; or (2) there was pre-clinical evidence to support efficient targeting of theaberrant function of the mutated gene and/or potentialclinical benefit; and (3) there was some age-appropriateinformation on dosing. Targetable somatic mutationswere further categorized using a five-tier system previ-ously described by Wagle et al. [26] and Harris et al.[15]. This sub-tiering system uses the strength ofpreclinical and clinical data as evidence to support thepotential clinical benefit of targeting an altered genewith a specific therapeutic agent.

ResultsPatientsDemographic and clinical characteristics are presentedin Table 1 and Fig. 1. Molecular characterization wasperformed on 120 samples (85, primary disease; 35,relapse/refractory disease) from 101 consecutive cases(mean age, 9.3 years; median age, 8.0 years; range, 2weeks – 26 years). Patients aged over 18 years in this co-hort were initially diagnosed with a pediatric diseaseunder the age of 18 years. Testing included: full cWES(tumor, germline, and transcriptome; n = 63); cWESwithout transcriptome (n = 19); transcriptome only (n =3); targeted tumor panel sequencing (n = 13); and

constitutional WES (proband and parental blood) (n =22). For constitutional WES, trios (proband and bothparents) were performed in 18/22 cases, 3/22 cases onlyhad one parent available for testing, and in one caseonly the proband was tested post-mortem. Eighty-fourpatients underwent single platform testing, while mul-tiple sequencing platforms were used for 17 cases (36samples). Cases were predominantly pediatric patientswith solid tumors (64%) (Fig. 1; Additional file 2: TableS3). Sarcoma (n = 17) was the most common diagnosticsub-category followed by brain tumors (n = 16). Patientswith lymphoid disease (n = 17) comprised the majorityof hematological conditions (Fig. 1; Additional file 2:Table S3).

Informed consent, cost, and reimbursementAll patients were consented to genomic analysis eitherthrough a research consent or clinical WES consent.Of the 101 cases, 67 were consented using the clinicalcWES consent. Only four (6%) opted out of learningsecondary findings and 21 (31%) opted out of having

Table 1 Patient and sample characteristics (n = 101)

n (%)

Diagnostic category

Solid tumors 65 (64)

Hematologic conditions 36 (36)

Gender

Male 60 (59)

Female 41 (41)

Age (years)

0–5 35 (34)

6–12 30 (30)

13–17 20 (20)

≥18 16 (16)

Samples tested (n = 120)

Primary disease 85 (71)

Relapse/refractory 35 (29)

Platform (n = 120)

Cancer WES with transcriptome 63 (53)

Cancer WES 19 (16)

Constitutional WES 22 (18)

Transcriptome only 3 (2)

Targeted panel (467 genes) 13 (11)

Normal tissue source (n = 104)

Blood 78 (74)

Buccal swab 23 (23)

Unaffected tissue 3 (3)

WES whole-exome sequencing

Oberg et al. Genome Medicine (2016) 8:133 Page 5 of 19

secondary findings in their medical record. All patientsconsented to have leftover samples stored. Only asingle patient (2%) opted out of future contact(Additional file 2: Table S4).As part of clinical implementation, we assessed the

cost of cWES and RNA-seq and the reimbursementlandscape. The total cost per case was calculated bysumming the total variable cost (reagent cost, patholo-gist time) with the fixed cost per case (annual machinecost, annual maintenance, tech labor cost, informaticscost, space for NGS hardware, server time, NGS analysislease, and data storage). The estimated cost of WES(tumor/normal) was $4459 and the cost of RNA-seq was$1764. These estimates do not include administrativeoverhead and billing for services.The time to receiving final reimbursement decisions

from third-party payers ranged between 6 months to1 year. To date, we have received a decision for 56patients with 45/56 (80%) receiving partial reimburse-ment. The average reimbursement by carrier type wasas follows: commercial, $2747 (range, $770–6917);managed government plans, $2918 (range, $750–4555); and $0 from government plans. Patients andtheir families were not charged for sequencing oranalysis.

Genomic alterations in pediatric solid tumors andhematologic disordersOver 150-fold and 500-fold average coverage wasachieved by WES and targeted capture sequencing,respectively with >98% of the coding sequences havingat least tenfold coverage. The mean mutational loadacross patients was 216.9 variants (SD = 829.3, median= 69), with a higher mean mutational load in solidtumors compared to hematologic malignancies (Fig. 2;Additional file 4: Figure S1). Genomic aberrations werereported in 92/101 patients (91%). After filtering, atotal of 180 mutations (Additional file 2: Table S5) and20 fusions were reported, 110 (including 10 fusions)from solid tumor samples (mean number of aberrationsper sample, 2.91; median, 2.00; range, 1–6) and 90 (in-cluding 10 fusions) from hematologic samples (meannumber of aberrations per sample, 5.2; median, 4.0;range, 1–12). The most commonly mutated gene wasTP53 (n = 9, 9%) in solid tumor samples and RAS path-way constituents (NRAS: n = 5, 5%; KRAS: n = 3, 3%) inhematologic samples (Fig. 3). In addition, significantchanges in the pattern of genetic alterations were notedon serial sequencing of samples from individual patientsat different time points during their therapy, reflectingclonal evolution. Awareness of these changes is important

Fig. 1 PIPseq overview. An overview of the PIPseq patients sequenced is presented on the left and a pie chart showing the distribution ofdiagnostic categories on the right

Oberg et al. Genome Medicine (2016) 8:133 Page 6 of 19

for selecting an appropriate targeted therapy and assessingresponse to therapy.

Evaluating potential clinical utility and targetable alterationsA genetic variant was considered targetable if: (1) anFDA-approved drug or experimental drug was availablethat inhibited the target directly or inhibited its down-stream signaling pathway; or 2) there was preclinicalevidence to support efficient targeting of the aberrantfunction of the mutated gene and/or potential clinicalbenefit; and 3) there was some age-appropriate infor-mation on dosing. Consistent with the published rec-ommendations from the Association for MolecularPathology [25], we evaluated clinical utility based on“the ability of a test result to provide information to thepatient, physician, and payer related to the care of thepatient and his/her family members to diagnose, moni-tor, prognosticate, or predict disease progression, andto inform treatment and reproductive decisions.”

Targetable somatic genomic alterationsOverall, 38/101 patients (38%) had at least one poten-tially targetable genomic alteration (Table 2). Specifically,21/65 patients (32%) with solid tumors and 17/36 (47%)patients with hematologic conditions carried targetablealterations. Matched therapy based on genomic findingswas received in 6/38 patients (16%).

Examples of targetable alterations include the iden-tification of a cKIT (p.Asn655Lys) [27] mutation in a7-year-old boy with acute myeloid leukemia (AML),who was subsequently treated with palliative imatiniband achieved a near-complete clearing of peripheralblood leukemic blasts with a sustained response for 9months. The RNA expression data also led us to iden-tify a BCR-ABL1-like [28] expression pattern in a 9-year-old girl with relapsed, refractory B-cell acutelymphoblastic leukemia (ALL). Subsequent analysisidentified a NUP214-ABL1 [29] fusion by real-timepolymerase chain reaction (RT-PCR) and the additionof dasatinib to the third-line induction regimen re-sulted in a deep remission allowing for a curative bonemarrow transplant. These results demonstrate the util-ity of comprehensive genomic characterization to iden-tify clinically actionable alterations in pediatriconcology patients.

Clinical impact of non-targetable somatic mutationsWhile many studies have focused on actionable alter-ations, the potential clinical impact of non-targetablealterations was also evaluated. Genomic alterationsidentified by sequencing helped to confer a moleculardiagnosis in 23 patients and identified prognostic,pharmacogenomic, and other significant health main-tenance recommendations in 32 patients (Table 3).Although these findings do not meet the definition of“actionability,” the clinical impact of such findings canbe quite profound. For example, the identification of aSTAT5B mutation [30] in a 5-year-old girl erroneouslydiagnosed with T-cell ALL helped to establish a diagno-sis of gamma-delta T-cell lymphoma. Also, identifica-tion of a PTPN11 mutation in a 4-year-old boycontributed to a change in his diagnosis from de novoAML to juvenile myelomonocytic leukemia (JMML)which evolved into AML [31].The identification of resistance alleles likewise is not

considered actionable, but may carry significant clinicalimplications. For example, in the 9-year-old girl with therelapsed NUP214-ABL1 B-ALL, the finding of a NT5C2mutation associated with resistance to nucleoside analogtherapies [32, 33] had clear implications for her salvagetherapy. In aggregate, sequencing results were clinicallyinformative for diagnostic, prognostic, or pharmacoge-nomic purposes in 38 patients (38%).

Clinical impact of transcriptome and CNV analysis beyondtarget identificationClinical impact by RNA-seq and CNV analysis was dem-onstrated in 23/33 patients (70%) (Table 3). Gene fusionsconfirming diagnosis were found in five patients: BCR-ABL1 (chronic myeloid leukemia), ASPSCR1-TFE3 (al-veolar soft part sarcoma), EWSR1-FLI1 in two patients

Fig. 2 Somatic mutation load by diagnostic category. Box plotscomparing overall somatic mutation rates across solid tumors andhematologic conditions detected by NGS. The top and bottomends of the boxes represent the 25th and 75th percentile values,respectively, and the segment in the middle is the median. Thetop and bottom extremes of the bars extend to the minimum andmaximum values. The box plot depicts the total mutation loadexcluding four outliers (one solid tumor and three hematologic).See Additional file 4: Figure S1 for inclusive dataset with outliers.The total mutational load (prior to filtering or orthogonal validation)for solid tumors was 4972 variants (mean, 84.3; SD, 43.9; median, 85;range, 15–214) and for hematologic conditions was 1478 variants(mean, 56.85; SD, 34.9; median, 47; range, 14–149)

Oberg et al. Genome Medicine (2016) 8:133 Page 7 of 19

(Ewing sarcoma), and EWSR1-WTI (desmoplastic smallround cell tumor). A novel EML4-NTRK3 fusion foundin a 2-year-old boy supported a change in diagnosisfrom undifferentiated sarcoma to infantile fibrosarcoma[15, 34]. In one patient, a CBFA2T3-GLIS2 [35] fusionconfirmed the diagnosis of acute megakaryoblasticleukemia (AMKL), was associated with poor prognosis,and supported the recommendation for a bone marrowtransplant. A PAX7-FOXO1 fusion was diagnostic andprognostic in a toddler with histologically defined solidalveolar rhabdomyosarcoma, but in whom FISH ana-lysis using the FOXO1A (FKHR; 13q14.1) break-apartprobe was repeatedly negative.CNV was inferred from the WES data and relative

gene expression was determined by reference to an aver-aged gene expression model. Segmental and gene

expression changes having prognostic implications wereidentified in 11 patients with a variety of diagnoses. Fourpatients diagnosed with neuroblastoma could be strati-fied based on RNA-seq and CNV: one high-risk patientwith MYCN amplification, LOH at 1p and 11q, gain of17q, and MYCN overexpression; one high-risk patientwith MYCN amplification, LOH at 1p, gain of 17q, andMYCN overexpression; one high-risk patient withoutMYCN amplification or LOH at 1p and 11q, and noevidence of MYCN overexpression; and one intermediate-risk patient without MYCN amplification or LOH at 1pand 11q and no evidence of MYCN overexpression.Medulloblastoma subgrouping was supported by over-expression and CNV in two patients. Poor prognosticfeatures were found in two other patients: low expressionof PAX8, FHIT, CASP10, CHD2, with high expression

Fig. 3 Summary of informative results from the PIPseq program. A matrix representation of findings with biological significance from thesequencing results are presented. Data are derived from all 101 patients that underwent WES of tumor-normal sample pairs, exome sequencingof germline DNA, transcriptome analysis of tumor, CNV of tumor, and targeted panel sequencing of tumor only. Deleterious mutations wereloss of function mutations and activating mutations refers to recurrent, previously reported activating mutations in oncogenes or variants withpublished in vitro evidence as being activating

Oberg et al. Genome Medicine (2016) 8:133 Page 8 of 19

Table 2 Sub-classification of potentially targetable somatic mutations for treatment planning

PIP ID Diagnosis Target alteration Mutation (change) Potential target therapy

Tier 1 (Data demonstrating benefit- same tumor type, same gene)

14-85546 CML BCR-ABL1 Fusion TKI

13-45348a AML IDH1 c.394C > T (p.R132C) IDH Inhibitor

14-24794a AML c-KIT; TET2; FLT3 c.2446G > C (p.D816H); c.3663delT(p.C1221Wfs); c.2505 T > G (p.D835E)

TKI; Hypomethylatingagent; TKI

14-53198 AML TET2 c.1156G > T (p.V386L) Hypomethylating agent

13-77086 AML c-KIT c.1965 T > G (p.Asn655Lys) TKIb

Tier 2 (Data demonstrating benefit- different tumor type, same gene)

15-18928 Hepatic rhabdoid tumor SMARCA4 c.3574C > T (p.R1192C) EZH2 Inhibitor

14-13487 Osteosarcoma TSC1 c.2503-1G > C (p.?) mTOR Inhibitor

14-47205 Nephroblastomatosis PIK3CA c.1035 T > A (p.N345K) PI3K/AKT/mTOR Inhibitor

15-40141 Pleomorphicxanthoastrocytoma

TMEM106B-BRAF Fusion MEK Inhibitor

Tier 3 (Published, presented, in press pre-clinical data demonstrating benefit- same tumor type, same gene)

15-64793 ALL FOXP1-ABL1 Fusion TKIb

15-26188 ALL NUP214-ABL1 Fusion TKIb

15-79700 ALL NRAS c.183A > T (p.Q61H) MEK Inhibitor

14-20062 ALL KRAS c.183A > T (p.Q61H) MEK Inhibitor

14-24794a AML NRAS c.183A > C (p.Q61H) MEK Inhibitor

15-29224 AML JAK3 c.1718C > T (p.A573V) JAK Inhibitor

13-45348a AML NRAS c.38G > C (p.G13A) MEK Inhibitor

13-95124 AML NRAS; MLL-AFF1 (KMT2A-AFF1) c.182A > G (p.Q61R); Fusion MEK Inhibitor;DOT1L Inhibitor

14-45760 AML NRAS c.38G > A (p.G13D) MEK Inhibitor

13-50662 AML PTPN11 c.1508G > T (p.G503V) MEK Inhibitor

14-15491 AML NUP98-NSD1 Fusion DOT1L Inhibitor

14-27243 Neuroblastoma NRAS c.181C > A (p.Q61K) MEK Inhibitor

14-70449 Rhabdomyosarcoma NRAS c.181C > A (p.Q61K) MEK Inhibitor

14-42817 Neuroblastoma KRAS c.34G > T (p.G12C) MEK Inhibitor

15-11925a Osteosarcoma MYC Overexpression BET Inhibitor

16-74654 Rhabdomyosarcoma FGFR4 c.1582G > T (p.G528C);c.1648G > C (p.V550L)

FGFR4 Inhibitor

15-23518 Rhabdomyosarcoma FGFR4 c.1648G > C (p.V550L); c.1949G > T(p.R650L); Overexpression

FGFR4 Inhibitor

14-37237 Glioblastoma multiforme Gain 12q.14.1 involving CDK2 Copy number CDK4/6 Inhibitor

15-44470 Medulloblastoma PTCH1, SUFU, ZIC3 Overexpression SMO Inhibitor

15-10838 Glioma H3F3A; FGFR1 c.83A > T (p.K28M);c.1731C > A (p.N577K)

HDAC Inhibitor

15-27992 Hepatic rhabdoid tumor Homozygous deletion of chr 22q11.23with homozygous deletion of SMARCB1;Loss of expression of SMARCB1

Copy number; Loss of expressionwith biallelic deletion

EZH2 Inhibitor

Tier 4 (Published, presented, in press pre-clinical data demonstrating benefit- different tumor type, same gene)

15-36388 AML MLL3 (KMT2C) c.2110G > T (p.E704X) BET Inhibitor

13-72282 T-ALL KRAS; JAK1; STAT5B c.40G > A (p.V14I); c.3076A > G(p.K1026E); c.2110A > C (p.I704L)

MEK Inhibitor; JAK Inhibitor;BCL2/BCLX-L Inhibitor

15-66870 Adrenocorticalcarcinoma

ALK c.3436C > A (p.Q1146K) ALK Inhibitorb

Oberg et al. Genome Medicine (2016) 8:133 Page 9 of 19

of CHD11, FUS, and MTA1 in a patient with Ewingsarcoma [36], and gain of 1q and loss of 6q and overex-pression of TNC, CALB1, PLAG1, ALDH1L1, andRELN in a patient with ependymoma [37]. Overexpres-sion of CCND1 in a patient with hepatoblastoma wasconsidered a good prognostic indicator. One patientwith AML with a CBFB-MYH11 fusion could beassigned to risk-based therapy and the diagnosis ofgamma-delta T-cell lymphoma [38, 39] was also corrob-orated by CNV with isochromosome 7q.

Clinically impactful germline alterationsA total of 90 patients had germline tissue sequenced.Cancer WES included germline analysis in 68/90 pa-tients. Tumor sequencing plus constitutional WES wasperformed in eight patients and 14 patients had onlygermline tissue sequenced for a variety of indicationsincluding clinical suspicion of cancer predisposition orof an underlying immunologic defect responsible for thedevelopment of lymphoma or hemophagocytic lympho-histiocytosis (HLH).Clinically impactful germline alterations (Table 4) were

found in 18/90 patients (20%): 11/57 patients with solidtumors (19%) and 7/33 patients with hematologic condi-tions (21%). In the solid tumor category, two alterationsin APC were diagnostic: one in a patient with hepato-blastoma and a family history consistent with familialadenomatous polyposis (FAP; p.R1114) and one associatedwith newly appreciated Gardner’s syndrome (p.E1554fs) ina 14-year-old boy with pilomatricomas and epidermoidcysts prior to his carcinoma diagnosis. Two variants inATM (p.R189K, p.K2756*) were found in a 16-year-oldboy with medulloblastoma inferring an increased risk fordeveloping other cancers. All were referred for geneticcounseling and consideration for future cancer screeningin the patient and family.

In patients with hematologic conditions, the incidence ofgermline alterations related to the primary diagnosis was ob-served in five patients (15%). A homozygous pathogenicvariant in C1QA (p.Gln208Ter) diagnostic of C1Q defi-ciency was identified in a 2-year-old girl with HLH. Ahomozygous pathogenic variant in PMS2 (p.S459X) diag-nostic of congenital mismatch repair deficiency was identi-fied in one patient with T-cell lymphoblastic lymphoma andconsanguineous parentage [40]. A likely pathogenic variantin XIAP (p.R443P) was identified in a 6-year-old girl withHLH, recurrent EBV infections, and suspected underlyingimmunodeficiency. Germline testing also revealed a hetero-zygous pathogenic splicing variant in RUNX1 (c.806-2A >G,r.Spl) in a patient with AML referred for transplantation forpersistent thrombocytopenia following chemotherapy [41].Both an HLA-matched sibling with borderline low plateletsand the father were found to carry the same variant. An un-related donor source was selected. A 2-month-old patienthospitalized for fulminant hemophagocytic syndrome was re-ferred for evaluation of presumed familial HLH and was con-sidered for hematopoietic stem cell transplantation. However,germline WES identified a pathogenic homozygous mutationinMLL2 (p.M3881Cfs*9) establishing the diagnosis of Kabukisyndrome [42] and familial HLH was ruled out due to thelack of alterations in any HLH-associated genes and subse-quently plans for a bone marrow transplant were averted.ACMG secondary findings were identified in six patients

(Table 4) and were returned to the families by clinical gen-etics. A germline BRCA1 mutation was discovered in an18-year-old boy with a rare hepatic tumor and a 17-year-old girl with ependymoma. A TP53 mutation was found ina 1-year-old girl with AML, a TNNT2 mutation associatedwith dilated cardiomyopathy was found in a 15-year-oldboy with osteosarcoma, a RYR1 mutation associated withmalignant hyperthermia was found in a 7-year-old girl withneuroblastoma, and a mutation in VHL was found in a2-year-old boy with ependymoma.

Table 2 Sub-classification of potentially targetable somatic mutations for treatment planning (Continued)

Tier 5 (Anything else the molecular tumor board thought was sufficient to qualify for treatment planning)

15-16072 ALL SMARCC2-PDGFRB Fusion TKI

14-75899 Neuroblastoma CDK4, MDM2 Overexpression NEPENTHE trial [NCT02780128]

15-11925a Osteosarcoma MCL1, CCNE1 Overexpression CDK4/6 Inhibitor

15-35162 Osteosarcoma CUL4A Overexpression NAE Inhibitor

14-71727 Osteosarcoma RAD51C c.24 T > G (p.F8L) PARP Inhibitor

15-83826 Osteosarcoma PDGFRA, KDR (VEGFR2) Overexpression MTKIb

13-21968 Congenital fibrosarcoma EML4-NTRK3 Fusion ALK Inhibitor

14-84044 Inflammatorymyofibroblastic tumor

VCAN-IL23R Fusion JAK Inhibitorb

aSame patientbTargeted therapy receivedALL acute lymphoblastic leukemia, AML acute myeloid leukemia, CML chronic myeloid leukemia

Oberg et al. Genome Medicine (2016) 8:133 Page 10 of 19

Table 3 Clinical utility beyond targetable somatic mutations

PIP ID Diagnosis Alteration Mutation (change) Clinical utility Implication

Sequence mutations

15-63375 AML changed to JMML PTPN11; SETB1 c.181G > T (p.D61Y);c.2602G > A (p.D868N)

Diagnostic [31, 46] JMML

13-72282a T-ALL STAT5B c.2110A > C (p.I704L) Diagnostic [30, 47] Gamma-delta T-celllymphoma

15-26188 ALL NT5C2 c.1219G > T (p.D407Y) Pharmacogenomic[32, 33]

Affects therapy

13-45348 AML IDH1 c.394C > T (p.R132C) Diagnostic [48, 49] Maffucci syndrome

14-53198 AML CEBPA c.939_940insAAG(p.K313_V314insK);c.326_327insC (p.P109fs)

Prognostic [50] Improved prognosis

15-10838 Glioma H3F3A c.83A > T (p.K28M) Prognostic [51] Poor prognosis

14-37237 Glioblastoma multiforme H3F3A c.83A > T (p.K28M) Prognostic [52, 53] GBM subgroup, K27

14-35585 Renal cell carcinoma VHL c.497 T > G (p.V166G) Diagnostic [54] Von Hippel Lindau

14-47205 Nephroblastomatosis PIK3CA c.1035 T > A (p.N345K) Diagnostic [55] Nephroblastomatosis

14-78154a Medulloblastoma KDM6A c.2989_2990dupAT(p.M997fs)

Prognostic [56] Risk stratification, group 4

14-75899 Neuroblastoma ATRX c.5239delA (p.T1747fs) Prognostic [57] Poor prognosis

14-10141 Pleuropulmonaryblastoma DICER1 c.5438A > G (p.E1813G) Health Maintenance[58]

DICER syndrome

Transcriptome analysis and CNV

14-24794 AML CBFB-MYH11 Fusion Prognostic [59] Low-risk stratification

15-64793 B-ALL FOXP1-ABL1 Fusion Prognostic [60] High-risk stratification

15-84578 AMKL CBFA2T3-GLIS2 Fusion Diagnostic [61];Prognostic [35]

AMKL; Poor prognosis

14-85546 CML BCR-ABL1 Fusion Diagnostic [59] CML

13-72282a T-ALL Isochromosome 7q Copy number change Diagnostic [39] Gamma-deltaT-cell lymphoma

15-46387 Rhabdomyosarcoma PAX7-FOXO1 Fusion Diagnostic [62, 63];Prognostic [64]

Rhabdomyosarcoma;High-risk group

13-81192 Alveolar soft part sarcoma ASPSCR1-TFE3 Fusion Diagnostic [65] Alveolar soft part sarcoma

13-65217 Ewing sarcoma EWSR1-FLI1 Fusion Diagnostic [66] Ewing sarcoma

15-47087 Ewing sarcoma EWSR1-FLI1; Low expressionof PAX8, FHIT, CASP10, CHD2,with high expression ofCHD11, FUS, and MTA1

Fusion; Expressionpattern

Diagnostic [66];Prognostic [36]

Ewing sarcoma;Poor prognosis

13-21968 Undifferentiated sarcoma EML4-NTRK3 Fusion Diagnostic [34] Infantile fibrosarcoma

16-88073 Ependymoma C11orf95-RELA; Alternatinggains and losses on chr 11and 22, consistent with a“chromothripsis-like” pattern

Fusion; Copy numberchange

Prognostic [67];Diagnostic [68]

Poor prognosis;RELA-type supratentorialependymoma

14-27243 Neuroblastoma MYCN amplified, deletionat 1p and 11q, gain 17q;MYCN over expressed

Copy number change;Overexpression

Prognostic [69, 70] Risk-based therapy

14-42817 Neuroblastoma MYCN non-amplified,no LOH at 1p11q; MYCNnot over expressed

Copy number change;No overexpression

Prognostic [69, 70] Risk-based therapy

15-39486 Neuroblastoma MYCN non-amplified, noLOH at 1p11q; MYCNnot over expressed

Copy number change;No overexpression

Prognostic [69, 70] Risk-based therapy

14-44070 Neuroblastoma MYCN amplified, loss of 1p,gain of 1q and 17q;MYCN over expressed

Copy number change;Overexpression

Prognostic [69, 70] Risk-based therapy

Oberg et al. Genome Medicine (2016) 8:133 Page 11 of 19

Germline variants classified as VOUS (Additional file 5:Table S6) were not returned to patients except if they metthe following criteria: (1) the variant was predicted to bedestructive; (2) the variant was in a well validated cancer-associated gene; and (3) a second somatic alteration wasidentified or the variant was reduced to homozygosity inthe tumor. Clinical genetics returned a VOUS to fourpatient families meeting these criteria, including an ITK(p.V175V) mutation in a 7-year-old girl with Hodgkinlymphoma and Epstein-Barr virus, an SDHC (p.G75D)mutation was found in a 12-year-old boy with ALL, aDICER1 (p.D609Y) mutation in an 18-year-old boy withALCL, and an APC (p.V1822D) mutation in a 7-year-oldboy with Ewing sarcoma.

Clinical impact of WESTo determine the overall clinical impact of NGS canceranalysis, we evaluated each case as to whether thesequencing data were of potential utility to the referringphysician in a clinically meaningful manner. Overall, clin-ically impactful results were found in 67/101 cases (66%)(Fig. 4). Potentially actionable alterations were found in38% of cases. In 23% of the cases, the data obtained pro-vided diagnostic significance. Importantly, germline pre-disposition to cancer was identified in 14% of all cases.WES and RNA-seq allows for significant additional

analytical endpoints (CNV, fusions, gene expression)over targeted gene panels. Focusing on the 60 caseswith full tumor/normal WES and RNA-seq (cWES),

the resulting data were clinically impactful in 45 cases(75%) (Fig. 5). A total of 72 potentially clinically impactfulresults were found with cWES accounting for 85% of thefindings (tumor/normal WES: 45%, n = 32; RNA-seq: 40%,n = 29) followed by CNV (7%, n = 5) and RNA-seq andCNV together in 8% (n = 6). Of the 30 potentially target-able aberrations found, 14 were by tumor/normal WES,15 by RNA-seq, and one by CNV (Fig. 5).

DiscussionIn this report, we reviewed the results of the first 101patients evaluated in our precision cancer medicineprogram. While we used a variety of analytical approachesmatched to the clinical indications, we primarily utilized acombination of tumor/normal WES and tumor RNA-seq.This platform provided several advantages over targetedcancer gene panels, including the ability to identify trans-locations, segmental chromosomal changes, and relativegene expression changes.Similar to other sequencing efforts in pediatric oncology,

we found that the overall mutational load in our patientswas relatively low by comparison to adult cancers [38]. Ofsignificance, we identified germline alterations that predis-pose to cancer in 14% of our patients. This is slightly higherthan other studies that have demonstrated approximately8.5–10% frequency of germline risk alleles in pediatriconcology patients and may reflect a selection bias to se-quence patients with high-risk cancers [12, 43, 44]. These

Table 3 Clinical utility beyond targetable somatic mutations (Continued)

15-88980 Hepatoblastoma Amplification of 11q13.2including CCND1; Overexpression of CCND1

Copy number change;Overexpression

Prognostic [71] Good prognosis

15-49177 Medulloblastoma IMPG2, GABRA5, LAPTM4B,MAB21L2, NPR3, MFAP4, NRL,ZFPM2, TSHZ3, IGF2BP3,GALNT14, GPR98; Loss of10q22.2-10qter involvingPTEN and SUFU, loss of17p, gain of 17q

Overexpression;Copy number change

Prognostic [72] Risk stratification,subgroup 3/4

15-70532 Ependymoma Gain of 1q, loss of 6q Copy number change Prognostic [73–75] Poor prognosis

14-78154a Medulloblastoma KCNA1, RBM24, KLHL13, EN2,SNCAIP, PDE1C, GRM8, KCNIP4,EXPH5, UNC5D, NID2, ST18,GPR12, SH3GL3; i17q

Overexpression;Copy number change

Prognostic [72] Risk stratification,subgroup 4

15-40141 Pleomorphicxanthoastrocytoma

Gain of chromosome 7,LOH at 9p

Copy number change Diagnostic [76] Pleomorphicxanthoastrocytoma

15-97336 Small round blue cell tumor EWSR1-WT1 Fusion Diagnostic [77] DSRCT

15-34296 Ependymoma TNC, CALB1, PLAG1,ALDH1L1, RELN

Overexpression Prognostic [37] Risk stratification, group A,poor prognosis

15-80972 ATRT LOH at 22q11.21qter,including SMARCB1; ASCL1

Copy number change;Overexpression

Diagnostic [78];Prognostic [79]

ATRT; Improved prognosis

aSame patientALL acute lymphoblastic leukemia, AMKL acute megakaryoblastic leukemia, AML acute myeloid leukemia, ATRT atypical teratoid rhaboid tumor, CML chronicmyeloid leukemia, CNV copy number variation, JMML juvenile myelomonocytic leukemia

Oberg et al. Genome Medicine (2016) 8:133 Page 12 of 19

Table

4Clinicallyim

pactfulg

ermlinemutations

PIPID

Diagn

osis

Alteratio

nMutation

Chang

eMutationtype

Clinicalutility

Implication

Previously

know

nor

suspected

Hem

atolog

iccond

ition

s

14-19751

HLH

MLL2

c.11640_11640d

elG;

c.15631G

>A

p.M3881Cfs*9

(deno

vo);p.E5211K

Com

poun

dhe

terozygo

us,

frameshift

Diagn

ostic

Kabu

kisynd

rome;

Transplantationwith

held

No

15-90485

HLH

C1QA

c.622C

>T

p.Gln208Ter

Hom

ozygou

s,no

nsen

seDiagn

ostic

C1Q

deficiency

No

15-33031

MDS

GATA2

(deno

vo)

c.16de

lGp.(E6fs)

Heterozygou

s,frameshift

Supp

ortstransplant

recommen

datio

n;Gen

etic

coun

selingforfamily

Increasedriskforde

veloping

AMLandMDS

No

14-92247

ALL

PMS2

c.1376C>G

p.S459X

Hom

ozygou

s,missense

Diagn

ostic;H

ealth

mainten

ance/Gen

etic

coun

seling

Con

stitu

tionalm

ismatch

repairde

ficiencysynd

rome;

Lynchsynd

rome(paren

ts)

No

15-46877

HLH

XIAP

c.1328G>C

p.R443P

Missense

Diagn

ostic

X-linkedlymph

oproliferative

synd

rome2(XLP2);

Transplantationrecommen

ded

No

14-19750

AML

RUNX1

c.806-2A

>G

r.Spl?

Heterozygou

ssplicesite

Diagn

ostic

Familialplatelet

disorder;

Transplantationdo

norchange

dfro

msiblingto

unrelateddo

nor

No

Solid

tumors

14-56374

Hep

atob

lastom

aAPC

c.3340C>T

p.R1114

Non

sense

Diagn

ostic;H

ealth

mainten

ance/Gen

etic

coun

seling

Familialaden

omatou

spo

lypo

sis

Yes

15-33544

Poorlydifferentiated

carcinom

awith

focal

neuroe

ndocrin

edifferentiatio

n

APC

(deno

vo)

c.4660_4661insA

p.E1554fs

Fram

eshift

Diagn

ostic;H

ealth

mainten

ance/Gen

etic

coun

seling

Gardn

ersynd

rome;Familial

aden

omatou

spo

lypo

sis

No

15-35162

Osteo

sarcom

aRB1

c.1216-3A>G

p.?

Splicesite

Health

mainten

ance/

Gen

eticcoun

seling

Increasedriskforde

veloping

othe

rcancers

Yes

15-44470

Med

ulloblastoma

ATM

c.566G

>A;c.8266A

>T

p.R189K;p.K2756*

Missense,

nonsen

seHealth

mainten

ance/

Gen

eticcoun

seling

Increasedriskforde

veloping

othe

rcancers

No

15-78886

aPineob

lastom

aUGT1A1

*28allele(“(TA

)7TA

A”)

Hom

ozygou

sPh

armacog

enom

icDrugsensitivity

No

15-78886

aPineob

lastom

aDICER1

c.4807du

pCp.L1603Pfs

Fram

eshift

Health

mainten

ance/

Gen

eticcoun

seling

Risk

ofovarianSertoli-Leydig

celltumor

No

15-17264

Hep

atocellular

carcinom

aUGT1A1

*28allele

Heterozygou

sPh

armacog

enom

icDrugsensitivity

No

ACMGsecond

aryfinding

s

15-29224

AML

TP53

c.644G

>A

p.S215N

Missense

Affectstherapy;Health

mainten

ance/Gen

etic

coun

seling

Explains

lack

ofrespon

seto

conven

tionalthe

rapy;Increased

riskforde

veloping

othercancers

No

Oberg et al. Genome Medicine (2016) 8:133 Page 13 of 19

Table

4Clinicallyim

pactfulg

ermlinemutations

(Con

tinued)

14-59462

Nestedstromal

epith

elialtum

orof

theliver

BRCA

1c.68._69de

lAG

p.Glu23Valfs

Fram

eshift

Health

mainten

ance/

Gen

eticcoun

seling

Breastcancer

No

16-88073

Epen

dymom

aBRCA

1c.5587_5594d

elGTA

GCACT

p.V1863Lfs*35

Fram

eshift

Health

mainten

ance/

Gen

eticcoun

seling

Breastcancer

No

14-75899

Neuroblastoma

RYR1

c.6838G>A

p.V2280I

Missense

Health

mainten

ance/

Gen

eticcoun

seling

Malignant

hype

rthe

rmia

No

14-13487

Osteo

sarcom

aTN

NT2

c.422G

>A

p.Arg141G

lnHeterozygou

sHealth

mainten

ance/

Gen

eticcoun

seling

Dilatedcardiomyopathy

No

15-34296

Epen

dymom

aVH

Lc.539T>C

p.I180T

Missense

Health

mainten

ance/

Gen

eticcoun

seling

VonHippe

lLindausynd

rome

No

a Sam

epa

tient

ALL

acutelymph

oblasticleuk

emia,A

MKL

acutemeg

akaryo

blastic

leuk

emia,A

MLacutemyeloid

leuk

emia,H

LHhe

mop

hago

cytic

lymph

ohistio

cytosis,JM

MLjuvenile

myelomon

ocyticleuk

emia

Oberg et al. Genome Medicine (2016) 8:133 Page 14 of 19

results underscore the need to routinely incorporate germ-line analysis for pediatric oncology patients.Although there is a paucity of Tier 1 actionable

alterations in pediatric cancers, using a more lenientdefinition of actionable which includes same gene–different tumor type, likely pathogenic VOUS, andassessment of both clinical and preclinical data,resulted in the identification of potentially actionablealterations in 38% of all patients. This is comparableto other studies and may in itself be sufficient justifi-cation for comprehensive genomic analysis in cancerpatients [10, 12, 15, 45]. Despite this finding, only16% of patients received matched targeted therapy.The ability to intervene with targeted therapies isparticularly challenging for pediatric patients. Manynewer drugs lack efficacy data in pediatric diseases orsafety data in children and are therefore not yetapproved for administration. Additionally, insurancecompanies are not obligated to provide coverage forthe off-label use of these high-cost agents. Compas-sionate use experimental therapies undergoing clinicaltesting or recently approved agents for adults are alsorarely granted for pediatric patients. Finally, a numberof targeted agents are not anticipated to have single-agent efficacy (e.g. MEK inhibition for RAS mutanttumors). Together, the lack of pediatric experience

and opportunities with combination therapy representadditional constraints in pediatric oncology.Nevertheless, we believe that narrowing the definition of

benefit to the identification of actionable targets andmatched targeted therapy underestimates the potentialclinical utility of comprehensive genomic analysis. Weprovide examples of genomic alterations that are notactionable per se, but which have significant clinical impactincluding for diagnostic, prognostic, or pharmacogenomicspurposes. Taking a broad view of clinical impact, it is not-able that the data from our sequencing platform impactedclinical decision-making in over two-thirds of all cases.With the increase in genomic medicine programs and thegrowing body of knowledge, the adoption of a more inclu-sive definition of clinical utility that does not narrowly focuson drug selection for patients with a specific biomarker isan important point to consider when incorporating NGStechnologies into clinical practice.Most cancer sequencing programs focus on interro-

gation of tumor DNA. It is notable that in our pro-gram the transcriptome data were responsible for anumber of clinically impactful calls that were not evi-dent from interrogating the DNA alone. In additionto verifying variants identified in the DNA analysis,the transcriptome was used to identify translocationsand was mined to identify signaling pathway activity.

Fig. 4 Clinically impactful results. The PIPseq experience yielded clinically impactful results in 67/101 cases. The Venn diagrams depict thecomplexity of overlapping findings within patients. That is, a patient may have a single finding fitting more than one category, whereasanother patient may have a finding fitting one category and another finding fitting a different category. For example, results categorized asTargetable/ Diagnostic (n = 6) are as follows: BCR-ABL1; IDH1; PIK3CA; EML4-NTRK3; [STAT5B, KRAS, JAK1/ STAT5B, i7q]; and [TMEM106B-BRAF/ gainchr 7, LOH 9p], with non-bracketed results representing a single finding fitting two categories and results within brackets representing those thatwere Targetable/ Diagnostic, respectively. Similarly, results categorized as Targetable/ Prognostic (n = 7) are as follows: FOXP1-ABL1; [TET2/ CEBPA];[H3F3A, FGFR1/ H3F3A]; [NRAS/ MYCN amp, del 1p and 11q, gain 17q]; [c-KIT, TET2, FLT3, NRAS/ CBFB-MYH11]; [KRAS/ No LOH 1p11q]; and[Gain 12q.14.1 involving CDK2/ H3F3A]. Individual patient results are provided in Tables 2, 3, and 4

Oberg et al. Genome Medicine (2016) 8:133 Page 15 of 19

We generated a model from transcriptomes in ourdatabase, allowing us to identify expression outliers.We were also able to project the gene expression datainto existing gene expression datasets for classificationpurposes, allowing us for example, to identify a BCR-ABL1-like gene expression pattern. Therefore, asses-sing tumor RNA is an important component of com-prehensive genomic approaches and in our seriessamples interrogated with both WES and RNA-seqcharacterization resulted in clinically impactful data in75% of cases.The importance of assessing germline in addition to

cancer DNA is evident from the 14% incidence of germ-line variants that may predispose to cancer. These find-ings clearly have broad implications that impact not onlythe patient but potentially the entire family. Moreover,the identification of germline risk offers opportunitiesfor prevention and early screening and detection. It is

notable that given the opportunity to opt out of thisknowledge, nearly all families actively choose for thereturn of these results, underscoring the fallacy of thepaternalistic view that families need to be protectedfrom learning these findings.Finally, extending beyond a fuller appreciation for

the potential clinical impact of sequencing technolo-gies, it is important to consider that genomic ap-proaches do not just provide incremental data, butmay replace many conventional tests. Currently, manygenetic alterations can be identified by standard ap-proaches, such as karyotype and FISH, and with fasterturnaround times. Similarly, existing NGS panels,which allow the detection of mutations and/or fusionsof clear clinical relevance, may be adequate in certainclinical scenarios. Nevertheless, in an era where initialdiagnostic biopsies are often performed through min-imally invasive approaches, there is a compelling

Fig. 5 Clinical impact of WES and RNA-seq by sequencing technology. Sixty patients had full tumor/normal WES (including CNV) and RNA-seq(cWES) performed. A total of 72 clinically impactful results were found in 45/60 cases (75%). A pie chart of the overall clinical impact ofcWES is presented on the left with a pie chart and table showing the number of impactful findings by sequencing technology on the right.For six patients, CNV and overexpression together yielded prognostic information in four patients with neuroblastoma and two patientswith medulloblastoma

Oberg et al. Genome Medicine (2016) 8:133 Page 16 of 19

argument to utilize comprehensive approaches withminimal tissue requirements. As the cost of NGS de-clines, the ability to comprehensively interrogate thegenome may supersede the need for sequential, po-tentially tissue-exhausting directed testing, with theadded benefit of uncovering rare targetable and po-tentially unexpected genomic drivers.

ConclusionsOur results demonstrate the feasibility of incorporatingclinical NGS into pediatric hematology-oncology prac-tice. While the frequency of finding actionable alter-ations is consistent with reports of other pediatriconcology sequencing endeavors [10, 12, 15, 45], we feelthis singular attribute grossly underestimates the poten-tial clinical utility of these data. The ability to avoid in-effective/inappropriate therapies, to solidify a definitivediagnosis, and to identify pharmacogenomics modifiersall have clinical impact. Taking this more inclusiveview, it is striking that the sequencing data were foundto be clinically impactful in 66% of all cases testedthrough our program and in 75% of cases comprehen-sively assessed using cWES and RNA-seq. The valueproposition for next generation diagnostics, therefore,should be measured both on the clinical impact of thedata and the ability to replace multiple conventionalsingle endpoint assays with a single comprehensiveview of the genome.

Additional files

Additional file 1: A PDF file containing Supplementary Methods.Supplementary Methods DNA and RNA extraction; somatic variantcalling strategy; CNV; transcriptome analysis. (PDF 82 kb)

Additional file 2: A PDF file containing Supplementary Tables S1–S5.Table S1 Quality statistics for WES. Table S2 Accession ID for all genesand fusions referenced in the manuscript. Table S3 Individual diagnoses.Table S4 Cancer WES consent preferences. Table S5 Reported somaticalterations by tier and germline alterations by category. (PDF 240 kb)

Additional file 3: De-identified clinical cWES report. (PDF 157 kb)

Additional file 4: A TIFF file containing Supplementary Figure S1.Figure S1 Total somatic mutation load across all patients. The totalmutational load across patients was 19,308 variants (mean, 216.9; SD,829.3; median, 69). By diagnostic category, the total mutational load forsolid tumors was 5853 variants (mean, 97.5; SD, 111.7; median, 86; range,15–881), and for hematologic conditions was 13,455 variants (mean,463.9; SD, 1428.8; median, 52; range, 14–5950). (TIF 845 kb)

Additional file 5: An excel file containing Supplementary Table S6.Table S6 Somatic and germline variants of unknown significance.(XLSX 20 kb)

AbbreviationsACMG: American College of Medical Genetics; CNV: copy number variation;cWES: cancer whole exome sequencing; EMR: electronic medical record;FFPE: formalin fixed paraffin embedded; HLA: human leukocyte antigen;HLH: hemophagocytic lymphohistiocytosis; VOUS: variants of uncertainsignificance; WES: whole exome sequencing

AcknowledgementsThe authors are grateful to the patients, their families, and to the doctorsand nurses at Columbia University Medical Center.

FundingThe present work was supported by the Sohn Conference Foundation(ALK, JGB), Hyundai Hope on Wheels (JAO, JGB, ALK), Alex’s LemonadeStand Foundation (JGB, RJZ, DD), Pediatric Cancer Foundation (JGB, JAO),Taybandz/ Conquering Kidz Cancer (DJY, JGB), Jamie Deutsch Foundation(JGB, FDC), and Hope and Heroes (JAO, JGB, ALK). We would also like tothank and acknowledge funding and resources from the Herbert IrvingComprehensive Cancer Center.

Availability of data and materialsThe dataset supporting the conclusions of this article can be obtained fromthe cBioPortal for Cancer Genomics (http://cbioportal.org) by selecting theDATA SETS tab and clicking on the citation listed in the Reference column.In accordance with CUMC IRB approval, vcf files, translocations, and geneexpression data are available through the cBioPortal.

Authors’ contributionsJAO, JGB, MLS, and ALK conceived, designed, and supervised all phases ofthe project. DP, ANS, and RZ provided coordination of clinical samples andtesting. MMM and PLN directed the clinical laboratory and oversaw samplepreparation, data processing, and computational analysis. HH and HRprovided interpretation of clinical analysis. MMM, SJH, and AT carried outclinical analysis, interpretation, and reporting to the EMR. JGB, MLS, FSDC,JHG, DJY, and ALK provided clinical expertise and interpretation of clinicalsequencing data at the multi-disciplinary tumor board and were responsiblefor overseeing consent and return of results. WKC provided critical advice onthe return of germline findings and provided oversight of genetic counseling.CK provided genetic counseling. SJA and JP carried out data processingand computational analysis. DD provided bioinformatics support. JAO, JGB,SGE, and ALK provided funding acquisition. JAO, JGB, MLS, and ALK wrotethe paper. JAO, JGB, MLS, JHG, DY, FDC, DD, WKC, KC, ANS, MMM, and ALKrevised the paper. All authors contributed to data interpretation, discussions,and editing of the paper. All authors read and approved the final manuscript.

Competing interestsThe authors declare that they have no competing interests.

Consent for publicationThis retrospective review and publication of results from all consentingparticipants was approved by the CUMC IRB (IRB# AAAQ8170 and AAAP1200).

Ethics approval and consent to participateAll participants provided written informed consent for tumor and/orgermline sequencing analysis. Participants signed either a ColumbiaUniversity Medical Center (CUMC) Institutional Review Board (IRB)approved consent (IRB# AAAB7109 or AAAJ5811) or clinical consent.Our research was carried out in accordance with the Declaration ofHelsinki and Good Clinical Practice (GCP).

Author details1Department of Pediatrics, Columbia University Medical Center, New York, NY10032, USA. 2Department of Pathology and Cell Biology, Columbia UniversityMedical Center, New York, NY 10032, USA. 3Department of Clinical Genetics,Columbia University Medical Center, New York, NY 10032, USA. 4Departmentof Medicine, Columbia University Medical Center, New York, NY 10032, USA.5Department of Microbiology and Immunology, Columbia University MedicalCenter, New York, NY 10032, USA. 6Herbert Irving Comprehensive CancerCenter, Columbia University Medical Center, New York, NY 10032, USA.7Present address: Memorial Sloan Kettering Cancer Center, New York, NY10065, USA. 8Present address: MNG Laboratories, 5424 Glenridge Drive,Atlanta, GA 30342, USA.

Received: 1 July 2016 Accepted: 2 December 2016

Oberg et al. Genome Medicine (2016) 8:133 Page 17 of 19

References1. Murphy S, Xu J, Kochanek K. Deaths: Final Data for 2010. National vital statistics

reports, vol. 61 no 4. Hyattsville, MD: National Center for Health Statistics; 2013.2. Oeffinger KC, Mertens AC, Sklar CA, Kawashima T, Hudson MM, Meadows AT,

et al. Chronic health conditions in adult survivors of childhood cancer.N Engl J Med. 2006;355(15):1572–82.

3. Hunger SP, Mullighan CG. Acute lymphoblastic leukemia in children. N Engl J Med.2015;373(16):1541–52.

4. Pinto NR, Applebaum MA, Volchenboum SL, Matthay KK, London WB,Ambros PF, et al. Advances in risk classification and treatment strategiesfor neuroblastoma. J Clin Oncol. 2015;33(27):3008–17.

5. Beltran H, Eng K, Mosquera J, Sigaras A, Romanel A, Rennert H, et al.Whole-exome sequencing of metastatic cancer and biomarkers of treatmentresponse. JAMA Oncol. 2015;1(4):466–74.

6. Frampton GM, Fichtenholtz A, Otto GA, Wang K, Downing SR, He J,et al. Development and validation of a clinical cancer genomicprofiling test based on massively parallel DNA sequencing. Nat Biotech.2013;31(11):1023–31.

7. Hovelson DH, McDaniel AS, Cani AK, Johnson B, Rhodes K, Williams PD, et al.Development and validation of a scalable next-generation sequencing system forassessing relevant somatic variants in solid tumors. Neoplasia. 2015;17(4):385–99.

8. Lanman R, Mortimer S, Zill O, Sebisanovic D, Lopez R, Blau S, et al.Analytical and clinical validation of a digital sequencing panel forquantitative, highly accurate evaluation of cell-free circulating tumorDNA. PLoS ONE. 2015;10(10):e0140712.

9. Meric-Bernstam F, Brusco L, Shaw K, Horombe C, Kopetz S, Davies MA, et al.Feasibility of large-scale genomic testing to facilitate enrollment ontogenomically matched clinical trials. J Clin Oncol. 2015;33(25):2753–62.

10. Mody RJ, Wu Y, Lonigro RJ, Cao X, Roychowdhury S, Vats P, et al.Integrative clinical sequencing in the management of refractory or relapsedcancer in youth. JAMA. 2015;314(9):913–25.

11. Ortiz MV, Kobos R, Walsh M, Slotkin EK, Roberts S, Berger MF, et al.Integrating genomics into clinical pediatric oncology using themolecular tumor board at the Memorial Sloan Kettering Cancer Center.Pediatr Blood Cancer. 2016;63(8):1368–74.

12. Parsons D, Roy A, Yang Y, Wang T, Scollon S, Bergstrom K, et al. Diagnosticyield of clinical tumor and germline whole-exome sequencing for childrenwith solid tumors. JAMA Oncol. 2016;2(5):616–24.

13. Shao D, Lin Y, Liu J, Wan L, Liu Z, Cheng S, et al. A targeted next-generationsequencing method for identifying clinically relevant mutation profiles inlung adenocarcinoma. Sci Rep. 2016;6:22338.

14. Uzilov AV, Ding W, Fink MY, Antipin Y, Brohl AS, Davis C, et al. Developmentand clinical application of an integrative genomic approach to personalizedcancer therapy. Genome Med. 2016;8(1):1–20.

15. Harris M, DuBois S, Glade Bender JL, Kim A, Crompton BD, Parker E, et al.Multicenter feasibility study of tumor molecular profiling to inform therapeuticdecisions in advanced pediatric solid tumors: The individualized cancertherapy (icat) study. JAMA Oncol. 2016;2(5):608–15.

16. ACMG Board of Directors. Points to consider in the clinical application ofgenomic sequencing. Genet Med. 2012;14(8):759–61.

17. Green RC, Berg JS, Grody WW, Kalia SS, Korf BR, Martin CL, et al. ACMGrecommendations for reporting of incidental findings in clinical exome andgenome sequencing. Genet Med. 2013;15(7):565–74.

18. Magi A, Tattini L, Cifola I, D’Aurizio R, Benelli M, Mangano E, et al. EXCAVATOR:detecting copy number variants from whole-exome sequencing data.Genome Biol. 2013;14(10):R120.

19. Trapnell C, Roberts A, Goff L, Pertea G, Kim D, Kelley DR, et al. Differentialgene and transcript expression analysis of RNA-seq experiments withTopHat and Cufflinks. Nat Protoc. 2012;7(3):562–78.

20. Trapnell C, Roberts A, Goff L, Pertea G, Kim D, Kelley DR, et al. Corrigendum:Differential gene and transcript expression analysis of RNA-seq experimentswith TopHat and Cufflinks. Nat Protoc. 2014;9(10):2513.

21. Ge H, Liu K, Juan T, Fang F, Newman M, Hoeck W. FusionMap: detectingfusion genes from next-generation sequencing data at base-pair resolution.Bioinformatics. 2011;27(14):1922–8.

22. Ramsköld D, Wang ET, Burge CB, Sandberg R. An abundance of ubiquitouslyexpressed genes revealed by tissue transcriptome sequence data. PLoSComput Biol. 2009;5(12):e1000598.

23. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, et al. The cBiocancer genomics portal: an open platform for exploring multidimensionalcancer genomics data. Cancer Discov. 2012;2(5):401–4.

24. Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, et al.Integrative analysis of complex cancer genomics and clinical profilesusing the cBioPortal. Sci Signal. 2013;6(269):l1.

25. Joseph L, Cankovic M, Caughron S, Chandra P, Emmadi R, Hagenkord J, et al.The spectrum of clinical utilities in molecular pathology testing procedures forinherited conditions and cancer. J Mol Diagn. 2016;18(5):605–19.

26. Wagle N, Berger MF, Davis MJ, Blumenstiel B, Defelice M, Pochanard P, et al.High-throughput detection of actionable genomic alterations in clinical tumorsamples by targeted, massively parallel sequencing. Cancer Discov. 2012;2(1):82–93.

27. Kinoshita K, Hirota S, Isozaki K, Nishitani A, Tsutsui S, Watabe K, et al.Characterization of tyrosine kinase I domain c-kit gene mutationAsn655Lys newly found in primary jejunal gastrointestinal stromal tumor.Am J Gastroenterol. 2007;102(5):1134–6.

28. Mullighan CG, Su X, Zhang J, Radtke I, Phillips LA, Miller CB, et al. Deletionof IKZF1 and prognosis in acute lymphoblastic leukemia. N Engl J Med.2009;360(5):470–80.

29. Roberts KG, Li Y, Payne-Turner D, Harvey RC, Yang YL, Pei D, et al. Targetablekinase-activating lesions in Ph-like acute lymphoblastic leukemia. N Engl J Med.2014;371(11):1005–15.

30. Küçük C, Jiang B, Hu X, Zhang W, Chan JK, Xiao W, et al. Activating mutations ofSTAT5B and STAT3 in lymphomas derived from γδ-T or NK cells. Nat Commun.2015;6:6025.

31. Tartaglia M, Niemeyer CM, Fragale A, Song X, Buechner J, Jung A, et al.Somatic mutations in PTPN11 in juvenile myelomonocytic leukemia,myelodysplastic syndromes and acute myeloid leukemia. Nat Genet.2003;34(2):148–50.

32. Meyer JA, Wang J, Hogan LE, Yang JJ, Dandekar S, Patel JP, et al. Relapse-specific mutations in NT5C2 in childhood acute lymphoblastic leukemia.Nat Genet. 2013;45(3):290–4.

33. Tzoneva G, Perez-Garcia A, Carpenter Z, Khiabanian H, Tosello V, Allegretta M,et al. Activating mutations in the NT5C2 nucleotidase gene drivechemotherapy resistance in relapsed ALL. Nat Med. 2013;19(3):368–71.

34. Tannenbaum-Dvir S, Glade Bender JL, Church AJ, Janeway KA, Harris MH,Manuskhani MM, et al. Characterization of a novel fusion gene EML4-NTRK3in a case of recurrent congenital fibrosarcoma. Cold Spring Harb Mol CaseStud. 2015;1(1):a000471.

35. Gruber Tanja A, Larson Gedman A, Zhang J, Koss CS, Marada S, Ta HQ,et al. An Inv(16)(p13.3q24.3)-encoded CBFA2T3-GLIS2 fusion proteindefines an aggressive subtype of pediatric acute megakaryoblastic leukemia.Cancer Cell. 2012;22(5):683–97.

36. Ohali A, Avigad S, Zaizov R, Ophir R, Horn-Saban S, Cohen IJ, et al.Prediction of high risk Ewing’s sarcoma by gene expression profiling.Oncogene. 2004;23(55):8997–9006.

37. Witt H, Mack SC, Ryzhova M, Bender S, Sill M, Isserlin R, et al. Delineationof two clinically and molecularly distinct subgroups of posterior fossaependymoma. Cancer Cell. 2011;20(2):143–57.

38. Alexandrov LB, Nik-Zainal S, Wedge DC, Aparicio SA, Behjati S, Biankin AV, et al.Signatures of mutational processes in human cancer. Nature. 2013;500(7463):415–21.

39. Tamaska J, Adam E, Kozma A, Gopcsa L, Andrikovics H, Tordai A, et al.Hepatosplenic γδ T-cell lymphoma with ring chromosome 7, anisochromosome 7q equivalent clonal chromosomal aberration.Virchows Arch. 2006;449(4):479–83.

40. Durno CA, Sherman PM, Aronson M, Malkin D, Hawkins C, Bakry D, et al.Phenotypic and genotypic characterisation of biallelic mismatch repairdeficiency (BMMR-D) syndrome. Eur J Cancer. 2015;51(8):977–83.

41. Song WJ, Sullivan MG, Legare RD, Hutchings S, Tan X, Kufrin D, et al.Haploinsufficiency of CBFA2 causes familial thrombocytopenia with propensityto develop acute myelogenous leukaemia. Nat Genet. 1999;23(2):166–75.

42. Miyake N, Koshimizu E, Okamoto N, Mizuno S, Ogata T, Nagai T, et al.MLL2 and KDM6A mutations in patients with Kabuki syndrome. Am J MedGenet A. 2013;161A(9):2234–43.

43. Knapke S, Nagarajan R, Correll J, Kent D, Burns K. Hereditary cancer riskassessment in a pediatric oncology follow-up clinic. Pediatr Blood Cancer.2012;58(1):85–9.

44. Zhang J, Walsh MF, Wu G, Edmonson MN, Gruber TA, Easton J, et al.Germline mutations in predisposition genes in pediatric cancer. N Engl J Med.2015;373(24):2336–46.

45. Chang W, Brohl AS, Patidar R, Sindiri S, Shern JF, Wei JS, et al. MultidimensionalClinOmics for precision therapy of children and adolescent young adults withrelapsed and refractory cancer: a report from the center for cancer research.Clin Cancer Res. 2016;22(15):3810–20.

Oberg et al. Genome Medicine (2016) 8:133 Page 18 of 19

46. Sakaguchi H, Okuno Y, Muramatsu H, Yoshida K, Shiraishi Y, Takahashi M,et al. Exome sequencing identifies secondary mutations of SETBP1 and JAK3in juvenile myelomonocytic leukemia. Nat Genet. 2013;45(8):937–41.

47. Nicolae A, Xi L, Pittaluga S, Abdullaev Z, Pack SD, Chen J, et al. FrequentSTAT5B mutations in [gamma][delta] hepatosplenic T-cell lymphomas.Leukemia. 2014;28(11):2244–8.

48. Amary MF, Damato S, Halai D, Eskandarpour M, Berisha F, Bonar F, et al.Ollier disease and Maffucci syndrome are caused by somatic mosaicmutations of IDH1 and IDH2. Nat Genet. 2011;43(12):1262–5.

49. Pansuriya TC, van Eijk R, d’Adamo P, van Ruler MA, Kuijjer ML, Oosting J,et al. Somatic mosaic IDH1 and IDH2 mutations are associated withenchondroma and spindle cell hemangioma in Ollier disease andMaffucci syndrome. Nat Genet. 2011;43(12):1256–61.

50. Ho PA, Alonzo TA, Gerbing RB, Pollard J, Stirewait DL, Hurwitz C, et al.Prevalence and prognostic implications of CEBPA mutations in pediatricacute myeloid leukemia (AML): a report from the Children’s Oncology Group.Blood. 2009;113(26):6558–66.

51. Sturm D, Witt H, Hovestadt V, Khuong-Quang DA, Jones DT, Konermann C,et al. Hotspot mutations in H3F3A and IDH1 define distinct epigenetic andbiological subgroups of glioblastoma. Cancer Cell. 2012;22(4):425–37.

52. Chan K-M, Fang D, Gan H, Hashizume R, Yu C, Schroeder M, et al.The histone H3.3K27M mutation in pediatric glioma reprograms H3K27methylation and gene expression. Genes Dev. 2013;27(9):985–90.

53. Khuong-Quang D-A, Buczkowicz P, Rakopoulos P, Liu XY, Fontebasso AM,Bouffet E, et al. K27M mutation in histone H3.3 defines clinically andbiologically distinct subgroups of pediatric diffuse intrinsic pontine gliomas.Acta Neuropathol. 2012;124(3):439–47.

54. Rao Q, Chen JY, Wang JD, Ma HH, Zhou HB, Lu ZF, et al. Renal cellcarcinoma in children and young adults: clinicopathological,immunohistochemical, and VHL gene analysis of 46 cases withfollow-up. Int J Surg Pathol. 2011;19(2):170–9.

55. Gripp K, Baker L, Kandula V, Conard K, Scavina M, Napoli JA, et al.Nephroblastomatosis or Wilms tumor in a fourth patient with a somaticPIK3CA mutation. Am J Med Genet. 2016;Part A 9999A:1–11.

56. Dubuc AM, Remke M, Korshunov A, Northcott PA, Zhan SH, Mendez-Lago M,et al. Aberrant patterns of H3K4 and H3K27 histone lysine methylation occuracross subgroups in medulloblastoma. Acta Neuropathol. 2013;125(3):373–84.

57. Cheung NV, Zhang J, Lu C, Parker M, Bahrami A, Tickoo SK, et al. Associationof age at diagnosis and genetic mutations in patients with neuroblastoma.JAMA. 2012;307(10):1062–71.

58. Brenneman M, Field A, Yang J, Williams G, Doros L, Rossi C, et al. Temporalorder of RNase IIIb and loss-of-function mutations during developmentdetermines phenotype in DICER1 syndrome: a unique variant of the two-hittumor suppression model. F1000Res. 2015;10(4):214.

59. Arber DA, Orazi A, Hasserjian R, Thiele J, Borowitz MJ, Le Beau MM, et al.The 2016 revision to the World Health Organization classification of myeloidneoplasms and acute leukemia. Blood. 2016;127(20):2391–405.

60. Ernst T, Score J, Deininger M, Hidalgo-Curtis C, Lackie P, Ershler WB, et al.Identification of FOXP1 and SNX2 as novel ABL1 fusion partners in acutelymphoblastic leukaemia. Br J Haematol. 2011;153(1):43–6.

61. Masetti R, Pigazzi M, Togni M, Astolfi A, Indio V, Manara E, et al.CBFA2T3-GLIS2 fusion transcript is a novel common feature in pediatric,cytogenetically normal AML, not restricted to FAB M7 subtype. Blood.2013;121(17):3469–72.

62. Laé M, Ahn EH, Mercado GE, Chuai S, Edgar M, Pawel BR, et al. Global geneexpression profiling of PAX-FKHR fusion-positive alveolar and PAX-FKHRfusion-negative embryonal rhabdomyosarcomas. J Pathol. 2007;212(2):143–51.

63. Williamson D, Missiaglia E, de Reyniès A, Pierron G, Thuille B, Palenzuela G,et al. Fusion gene–negative alveolar rhabdomyosarcoma is clinically andmolecularly indistinguishable from embryonal rhabdomyosarcoma.J Clin Oncol. 2010;28(13):2151–8.

64. Skapek SX, Anderson J, Barr FG, Bridge JA, Gastier-Foster JM, Parham DM,et al. PAX-FOXO1 fusion status drives unfavorable outcome for childrenwith rhabdomyosarcoma: a children’s oncology group report. PediatrBlood Cancer. 2013;60(9):1411–7.

65. Ladanyi M, Lui M, Antonescu C, Krause-Boehm A, Meindl A, Argani P, et al.The der(17)t(X;17)(p11;q25) of human alveolar soft part sarcoma fuses theTFE3 transcription factor gene to ASPL, a novel gene at 17q25. Oncogene.2001;20(1):48–57.

66. Delattre O, Zucman J, Plougastel B, Desmaze C, Melot T, Peter M, et al.Gene fusion with an ETS DNA-binding domain caused by chromosometranslocation in human tumours. Nature. 1992;359(6391):162–5.

67. Parker M, Mohankumar KM, Punchihewa C, Weinlich R, Dalton JD, Li Y,et al. C11orf95-RELA fusions drive oncogenic NF-[kgr]B signalling inependymoma. Nature. 2014;506(7489):451–5.

68. Pajtler Kristian W, Witt H, Sill M, Jones DT, Hovestadt V, Kratochwil F, et al.Molecular classification of ependymal tumors across all CNS compartments,histopathological grades, and age groups. Cancer Cell. 2015;27(5):728–43.

69. Bosse K, Maris J. Advances in the translational genomics of neuroblastoma:From improving risk stratification and revealing novel biology to identifyingactionable genomic alterations. Cancer. 2016;122(1):20–33.

70. Attiyeh EF, London WB, Mossé YP, Wang Q, Winter C, Khazi D, et al.Chromosome 1p and 11q deletions and outcome in neuroblastoma.N Engl J Med. 2005;353(21):2243–53.

71. López-Terrada D, Gunaratne PH, Adesina AM, Pulliam J, Hoang DM, Nguyen Y,et al. Histologic subtypes of hepatoblastoma are characterized by differentialcanonical Wnt and Notch pathway activation in DLK+ precursors. Hum Pathol.2009;40(6):783–94.

72. Northcott PA, Korshunov A, Witt H, Hielscher T, Eberhart CG, Mack S, et al.Medulloblastoma comprises four distinct molecular variants. J Clin Oncol.2011;29(11):1408–14.

73. Kilday J-P, Mitra B, Domerg C, Ward J, Andreiuolo F, Osteso-Ibanez T, et al.Copy number gain of 1q25 predicts poor progression-free survival forpediatric intracranial ependymomas and enables patient risk stratification:a prospective european clinical trial cohort analysis on behalf of the Children’sCancer Leukaemia Group (CCLG), Société Française d’Oncologie Pédiatrique(SFOP), and International Society for Pediatric Oncology (SIOP). Clin Cancer Res.2012;18(7):2001–11.

74. Olsen T, Gorunova L, Meling TR, Micci F, Scheie D, Due-Tonnessen B, et al.Genomic characterization of ependymomas reveals 6q loss as the mostcommon aberration. Oncol Rep. 2014;32(2):483–90.

75. Rousseau A, Idbaih A, Ducray F, Criniere E, Fevre-Montange M, Jouvet A,et al. Specific chromosomal imbalances as detected by array CGH inependymomas in association with tumor location, histological subtypeand grade. J Neurooncol. 2010;97(3):353–64.

76. Marko N, Weil R. The molecular biology of WHO Grade II gliomas. Neurosurg Focus.2013;34(2):E1.

77. Gerald WL, Rosai J, Ladanyi M. Characterization of the genomic breakpointand chimeric transcripts in the EWS-WT1 gene fusion of desmoplastic smallround cell tumor. Proc Natl Acad Sci U S A. 1995;92(4):1028–32.

78. Jackson EM, Sievert AJ, Gai X, Hakonarson H, Judkins AR, Tooke L, et al.Genomic analysis using high-density single nucleotide polymorphism-based oligonucleotide arrays and multiplex ligation-dependent probeamplification provides a comprehensive analysis of INI1/SMARCB1 inmalignant rhabdoid tumors. Clin Cancer Res. 2009;15(6):1923–30.

79. Torchia J, Picard D, Lafay-Cousin L, Hawkins CE, Kim SK, Letourneau L,et al. Molecular subgroups of atypical teratoid rhabdoid tumours inchildren: an integrated genomic and clinicopathological analysis.Lancet Oncol. 2015;16(5):569–82.

• We accept pre-submission inquiries

• Our selector tool helps you to find the most relevant journal

• We provide round the clock customer support

• Convenient online submission

• Thorough peer review

• Inclusion in PubMed and all major indexing services

• Maximum visibility for your research

Submit your manuscript atwww.biomedcentral.com/submit

Submit your next manuscript to BioMed Central and we will help you at every step:

Oberg et al. Genome Medicine (2016) 8:133 Page 19 of 19